Human metapneumovirus (HMPV) was first reported in 2001 and has quickly come to be recognized as a significant agent of respiratory tract disease worldwide, especially in the pediatric population, in immunocompromised individuals, and in the frail elderly. We are using recombinant DNA methods to characterize viral molecular biology and pathogenesis and to develop attenuated derivatives of HMPV for use as a live intranasal pediatric vaccine. HMPV is an enveloped virus with a genome that is a single negative-sense strand of RNA of approximately 13.3 kb. We previously determined the first complete HMPV genome sequences for each of the two genetic subgroups, A and B. HMPV encodes nine proteins: by analogy to human respiratory syncytial virus, its better-known relative, the HMPV proteins are: N, nucleoprotein;P, phosphoprotein;M, matrix protein;F, fusion protein;the M2-1 and M2-2 putative RNA synthesis factors;SH, small hydrophobic protein;G, attachment glycoprotein;and L, viral polymerase. We also previously developed a reverse genetic system for HMPV whereby complete infectious virus can be generated in cell culture entirely from cloned cDNAs. This provides a method for engineering the genome in pursuit of basic studies and for designing vaccines. We found that four viral genes could be deleted individually and in various combinations with little or no effect on viral replication in vitro, namely: G, SH, M2-1 and M2-2. Evaluation of the attenuation and immunogenicity of these viruses in hamsters and African green monkeys indicated that the del-G virus (with or without the additional deletion of the SH gene) and del-M2-2 virus are promising candidates to be live attenuated vaccines against HMPV. Additional vaccine candidates were generated by replacing the N or P open reading frame of HMPV with its counterpart from the closely related avian metapneumovirus (AMPV) subgroup C. We showed that AMPV is attenuated in primates due to a natural host range restriction, and it was hoped that the introduction of AMPV genes into HMPV would confer this attenuation phenotype. Evaluation in hamsters and African green monkeys showed that this indeed was the case, and the P-replacement virus (HMPV-Pa) in particular is a promising vaccine candidate whereas the HMPV-Na virus was insufficiently attenuated and would need importation of further attenuating mutations. HMPV-Pa also is an attractive vaccine candidate because it replicates 25-fold more efficiently than HMPV in Vero cells, a phenotype that also was observed with AMPV and will facilitate vaccine manufacture (since Vero cells are the substrate for vaccine manufacture). We created seed stocks of wild type HMPV and the HMPV-Pa and del-M2-2 viruses for clinical trials. The viruses were generated in our laboratory under conditions using Current Good Manufacturing Practices as a guide so as to be appropriate for products for human use. Using a contract facility, we have generated clinical trial material for the wild type virus and the HMPV-Pa derivative;the del-M2-2 virus is in progress. The wild type virus will be analyzed in clinical trials in seropositive adults to evaluate re-infection in HMPV-experienced individuals and to establish a benchmark for replication against which the attenuated derivatives can be judged. The attenuated vaccine candidates will be evaluated successively in Phase I clinical trials in seropositive adults, seropositive children, seronegative children, and HMPV-naive infants.